FULL PAPER
palladium loadings and temperatures (Table 1). We found
that the reduction is ideally carried out at room temperature
(258C) with a catalyst loading of 0.05 mol% palladium.
used in our protocol has a limited influence on the reaction
outcome (Table 1, entry 7).[7] Essentially, it sequesters Pd0
nanoparticles and prevents contamination of the solvent and
product by palladium residues. In other words, inaccuracy in
the charcoal weighing has no effect upon the reproducibility
of the reaction.
Table 1. Optimisation studies.[a]
With the optimised conditions in hand, we screened for
activity in the hydrogenation reactions of a variety of sub-
strates possessing an alkene or an alkyne functionality
(Table 2). Whatever the nature of the alkene, the hydroge-
nation reaction can be carried out at room temperature
(258C), under mild conditions, with a catalyst loading as low
as 0.025 mol%. Only the reduction of extremely hindered
cholesterol (15) to 5a-cholestan-3b-ol (16) required a higher
loading of palladium (1 mol%), but still at room tempera-
ture and under 1 atm of hydrogen (Table 2, entry 8). Al-
though a higher palladium loading was necessary for the re-
duction of cholesterol (15), a brief survey of the literature
for the use of palladium-based catalysts for this reaction
makes our protocol highly competitive.[8] Such mild condi-
tions allow the selective reduction of chalcone 17 without
significantly affecting the carbonyl group, which is usually
prone to reduction (Table 2, entry 9). Alkynes are also
easily reduced to the corresponding alkane at low palladium
loadings (Table 2, entries 10 and 11). To our surprise, yna-
mide 22[9] proved to be averse to hydrogenation and only
the corresponding enamide, 23, can be isolated at a loading
of 1.5 mol% palladium, as an inseparable mixture of iso-
mers (E/Z, ꢀ10:90). Similar behaviour has already been ob-
served with Lindlarꢂs catalyst.[10]
Entry
Temperature
[8C]
Pd loading
ACHTUNGTRNE[NUNG mol%]
Yield[b]
[%]
1
2
3
4
5
6
65
65
65
65
45
25
25
25
25
1
0.5
0.1
0.05
0.05
0.05
0.05
0.005
0
99
99
99
99
99
99[c]
99
18
3
7[d]
8
9
[a] Reaction conditions: stilbene (5 mmol), Pd
ACHTUNGRTEN(NUNG OAc)2, charcoal (90wt%/
Pd(OAc)2), MeOH (6.5 mL), H2 (1 atm, balloon), 12 h. [b] Yield of isolat-
ACHTUNGTRENNUNG
ed product. [c] Contamination of the solvent <0.2 ppb (see text).
[d] Charcoal was omitted from this reaction.
Working at higher temperatures does not allow lower cata-
lyst loadings. We adjusted the amount of support in order to
have a 10:90 (wt:wt) PdACHTNUGTRNEUNG(OAc)2/charcoal ratio, resulting in
an approximate 5 wt% loading of palladium metal. We have
also shown that the background reaction resulting from
traces of palladium on the magnetic stirring bar and/or
glassware is negligible (Table 1, entry 9). Inductively cou-
pled plasma mass spectrometry (ICP-MS) analyses show
very little contamination of the solvent by palladium resi-
dues, with levels as low as 0.2 ppb. This result indicates that
virtually all of the palladium species (>99.99%) have been
adsorbed onto the charcoal and confirms the quantity of pal-
ladium metal loaded onto the charcoal (ꢀ5 wt% Pd/C).
Our new protocol allows excellent reproducibility of re-
sults, since the quantity of palladium introduced can be ac-
curately controlled by using a home-made stock solution of
To evaluate the added value of our in situ prepared cata-
lyst, we evaluated (Table 3) the catalytic activity of various
commercially available Pd/C catalysts for the reduction of
cholesterol (15) to 5a-cholestan-3b-ol (16) at a Pd loading
of 1 mol%. As expected, dramatic differences in catalytic
activity are observed, depending upon the Pd source, giving
low to good conversions. However, to the credit of our pro-
tocol, none of the commercial Pd/C catalysts, in our hands,
were as active as our home-made catalyst. Moreover, it
should be noted that PdACTHNUTRGNE(UGN OAc)2 is one of the cheapest sour-
Pd
G
ces of Pd that competes favourably against most of the com-
mercial Pd/C catalysts. As a consequence, this protocol
would also be particularly convenient on a preparative scale
from both economic and safety perspectives (palladium ace-
tate in not pyrophoric).
Next, we investigated the hydrogenolysis of O-benzyl pro-
tecting groups by using a similar catalyst system (Table 4).
Benzyl ethers are commonly employed in synthetic chemis-
try as transient protecting groups for alcohols, carboxylic
acids and amines (through a carboxybenzyl (Cbz) group).
Although many reaction conditions for their cleavage have
been described in the literature with bases, Lewis acids and
oxidising or reducing agents under forced conditions, the
combination of hydrogen and a Pd/C catalyst has proved to
be the mildest and most effective system.[11] We were
pleased to find that the in situ prepared Pd/C catalyst was
able to smoothly catalyse the cleavage of the O-benzyl pro-
for reactions conducted at low loadings of palladium, espe-
cially on a laboratory scale. Although we screened various
batches of Pd
reproduced our results. This can be explained by the highly
stable nature of Pd(OAc)2, which does not decompose over
time. On the other hand, the use of low loadings of commer-
cially available Pd/C, on a laboratory scale (0.1–10 mmol)
requires the weighing of small quantities of the heterogene-
ous catalyst, which leads to irreproducible results. Indeed,
the quality of batches and precision of the weighing of small
quantities (0.1–5 mg), as well as the disparity in palladium
distribution over the charcoal, are all aspects that are ad-
dressed by our protocol (vide infra). In addition, from a
safety perspective, the launch of the reaction does not re-
quire any special handling, since, in contrast to Pd/C cata-
lysts, PdACHTUNGTRENNUNG(OAc)2 is not pyrophoric. The quantity of charcoal
Chem. Eur. J. 2010, 16, 12440 – 12445
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12441